There is increasing evidence from both animals and humans that broadband non-phase-locked responses in high gamma frequencies (~60-200 Hz) represent an electrophysiological signature of cortical processing. Recordings in monkeys have shown that they are tightly linked to neuronal firing and are correlated with mean population firing rates. Although the exact neural mechanisms underlying these responses are still under investigation, they offer an important metric of cortical processing at the level of population dynamics. Electrocorticographic (ECoG) recordings from subdural electrodes implanted for clinical purposes provide an unusual opportunity to study this activity and to use it to study cortical function in humans. The first two hypotheses to be tested are (1) that ECoG high-gamma responses are sufficiently robust and specific to study, in single subjects, the temporal dynamics of cortical representations/processes relevant to perceptual and language tasks, and (2) that repetition suppression of ECoG high-gamma responses is an electrophysiological correlate of neural priming that can be used to help differentiate the aforementioned neural representations/processes. In addition, we hypothesize (3) that time- and frequency-dependent measures of causal interactions between ECoG recording sites can elucidate the functional connectivity and dynamics of distributed cortical networks responsible for perceptual and linguistic functions. Finally, we hypothesize (4) that ECoG high gamma responses, in combination with other ECoG indices of cortical activation, can be used to map cortical function prior to surgery and that ECoG can be used with other functional mapping techniques to help prevent post-operative impairments in patients undergoing surgery for intractable epilepsy. Specific aims: (1) Test whether ECoG responses in high gamma or other frequencies can differentiate cortical representations/processes (e.g. tools vs. animals;phonological vs. lexical) in humans. (2) Test the effects of repetition priming on ECoG responses in high-gamma and other frequencies. (3) Test whether ECoG measures of event-related functional interactions correspond to the predicted functional connectivity and temporal dynamics of cortical networks supporting perception and language. (4) Test the correspondence between event-related ECoG responses and electrocortical stimulation mapping, and to compare their ability to predict postoperative functional impairments in patients undergoing epilepsy surgery. These studies will help guide current and future translational clinical efforts using ECoG (and EEG) for functional mapping and brain-computer interfaces. PUBLIC HEALTH RELEVANCE: The research plan will use a recently discovered EEG index of human brain activation to study the temporal sequence by which different brain areas are activated and interact with each other during basic perceptual and language tasks. This research is made possible by the participation of patients who have had EEG electrodes implanted in or on the surface of their brain for the surgical treatment of medically refractory epilepsy. Basic knowledge derived from this research is expected to facilitate future investigations of the brain mechanisms of human perception and language, as well as the development of new tools for functional brain mapping to avoid impairments following brain surgery for epilepsy.